In the US, injury is the leading cause of alcohol-related death and alcohol is the leading risk factor for injury. Injury is also the number one cause of death among adolescents in the United States. Across the nation, over 15,000 13-to-20 year olds die from injuries each year. Although adolescents are legally proscribed from consuming alcohol in the US, many still manage to do so and their consumption can be followed by an injury. This has led many policymakers to conclude that personal consumption is the primary mechanism by which alcohol enhances the risk of adolescent injury. However, we hypothesize that of potentially equal or greater influence on an adolescent's alcohol-related injury risk is their exposure to alcohol consumption by others around them. Many adolescents may risk injury simply by being in a family or a neighborhood environment where alcohol is present, regardless of their consumption -- being caught in the crossfire between intoxicated bar patrons, committing suicide in the absence of an alcoholic parent, or being hit by a drunk driver while walking to school. Using the entire City of Philadelphia as a research site, we intend to conduct a population-based case-control study to determine the risk of injury that exposure to alcohol creates for adolescents. As they occur, we will rapidly ascertain the injury times, locations, and alcohol exposure status for adolescent cases of homicide, suicide, and unintentional injury death between 13 and 20 years old. We will then use random-digit dialing and face-to-face interviews to survey adolescent controls from within Philadelphia as to their locations and alcohol exposures at the time of a matching case's injury. We will measure alcohol exposure in terms of individual, family, and neighborhood alcohol exposure. The proposed study will also uniquely make use of data from 17 local, state, and national sources as well as real-time surveys to identify cases of adolescent injury and measure personal, family, and neighborhood exposure to alcohol for both cases and their matched controls. In doing this, we specifically aim to answer the questions: (1) When considered separately, do baseline (crude) individual, family, or neighborhood exposures to alcohol increase the risk of fatal adolescent injury?; (2) When considered together, do individual, family, and neighborhood exposures to alcohol have a modifying (confounding) effect on each other's risks of fatal adolescent injury?; (3) When considered together, do individual, family, and neighborhood exposures to alcohol have a synergistic (interaction) effect on each other's risks of fatal adolescent injury? [unreadable] [unreadable] [unreadable]